OBJECTIVE Using polysomnography, the gold standard for sleep assessment, this study aimed to describe the objective prevalence of insomnia in the São Paulo, Brazil, Epidemiologic Sleep Study cohort of 1,101 adults (20-80 years old). METHODS Objective insomnia was defined by meeting 1 of the following criteria: sleep onset latency >30 minutes (sleep initiating insomnia), wake after sleep onset lasting >30 minutes (sleep maintenance insomnia), total sleep time <360 minutes and a terminal wakefulness >30 minutes (insomnia with too short duration of sleep or early morning awakening), or a combination of the previous quantitative criteria (mixed disorder). Using validated questionnaires based on Diagnostic and Statistical Manual of Mental Disorders, 4th edition (DSM-IV) criteria, subjective insomnia was categorized into 3 groups: good sleepers, insomnia symptoms, and DSM-IV insomnia. RESULTS A total of 1,042 subjects participated in the study (95% response rate). The prevalence of objective insomnia was 32%. The subjective prevalence of insomnia symptoms was 45%, and the subjective prevalence of DSM-IV insomnia was 15%. Sociodemographic factors were similar in both the objective insomnia and the DSM-IV insomnia groups. Age, but not psychiatric symptoms, was predictive of objective insomnia. The subjective criteria were not adequately sensitive (36%) to identify objective insomnia, but were adequately specific (77%) to rule out polysomnography noninsomnia. CONCLUSIONS The prevalence of objective insomnia assessed by polysomnography was higher than the prevalence of subjective insomnia according to DSM-IV-validated questionnaires. Clinical trials.gov ID: NCT00596713.